KP EOS Calculator: Guidance for EMR Implementation
Updated 5/6/21

Raw equations for the calculator
The raw equation for the risk prediction model is available in the supplement to the Puopolo paper.¹

There was a mistake in the original publication’s supplemental section which we would like to correct here:

Sepsis risk score (SRS) = 1/(1 + e-^βx)

βx = 47.8398 + (0.8680[tempimp]) - (6.9325[ga4mdlng]) + (0.0877 [ga4mdlng_sq]) + (1.2256[romimp]) - (1.0488[approptx1]) - (1.1861[approptx2]) + (0.5771[j_gbscar(+)]) + (0.0427[j_gbscar(u)])**

** J_gbscar(u) this variable was incorrectly printed in original paper as: - (0.0427[j_gbscar(u)])

Table 1: Variables for Neonatal EOS Calculator

Term	Definition	Coefficient, β
SRS	Sepsis risk score at birth based on maternal variables or “prior probability of EOS”
tempimp	Highest maternal intrapartum temperature (Value to 0.1˚F)	0.868
Romimp	Transformed rupture of membranes time (ROM time in hours + 0.05)0.2	1.2256
ga4mdlng	Gestational age (weeks and days)	-6.9325
ga4mdlng_sq	Gestational age squared (weeks and days)	0.0877
Approptx1	1 if GBS specific antibiotics are given ≥2 hours prior to deliver OR any antibiotics given 2-3.9 hours prior to delivery, otherwise 0	-1.0488
Approptx2	1 if Broad-spectrum antibiotics given ≥4 hours prior to delivery, otherwise 0	-1.1861
J_gbscar(+)	1 if GBS status is positive, otherwise 0	0.5771
J_gbscar(u)	1 if GBS status is unknown, otherwise 0	0.0427
Intercept	EOS incidence 0.3/1000 Live Births	40.0528
	EOS incidence 0.4/1000 Live Births	40.3415
	EOS incidence 0.5/1000 Live Births	40.5656
	EOS incidence 0.6/1000 Live Births	40.7489

SRS provides the raw risk of sepsis at birth based on inputted data. This risk at birth is used to determine recommendations for enhanced observation (i.e. q 4 hour vitals for 24 hours if risk at birth ≥ 1/1000).

The posterior probability of EOS (incorporating clinical presentation in the first 24 hours) or the “composite risk score” is determined once the infant’s clinical presentation has been categorized based on the following descriptions:

Table 2: Classification of Infant’s Clinical Presentation and Likelihood Ratios

Clinical Illness	Likelihood ratio 21.2
Persistent need for NCPAP/HFNC/mechanical ventilation Hemodynamic instability requiring vasoactive drugs Neonatal encephalopathy / Perinatal depression Seizure Apgar Score @ 5 minutes < 5 Need for supplemental O2 > 2 hours to maintain oxygen saturations > 90%
Equivocal	Likelihood ratio 5.0
Single persistent physiologic abnormality lasting > 4 hrs or two or more physiologic abnormalities lasting > 2 hours Tachycardia (HR > 160) Tachypnea (RR > 60) Temperature instability (> 100.4˚F or < 97.5˚F) Respiratory distress (grunting, flaring, or retracting) not requiring supplemental O2
Well Appearing	Likelihood ratio 0.41
No persistent physiologic abnormalities

To calculate individual posterior probability for EOS risk for each infant we used Bayes’ theorem: prior odds x Likelihood Ratio (LR) = posterior odds. The probability of EOS risk at birth is converted to odds to determine the prior odds (probability = odds/ [odds+1] ). The finding of interest is the infant’s evolving clinical presentation after birth. We used the conservative LR based on the upper limit of the 95% confidence interval.

Posterior Probability of EOS (Incorporating) Clinical Presentation in the first 24 hours)

This score is used in the online calculator to determine culture/antibiotics recommendations. We have included a table, below, which shows the stratification of the composite risk score and the clinical recommendations:

Table 3: Stratification of Composite Risk Score and Clinical Recommendation

See below for a spreadsheet we have given people to follow our calculations. There may be slight differences in the spreadsheet and the web calculator based upon rounding at various stages. See the cited background article on choices in our implementation.²

References

1. Puopolo KM, Draper D, Wi S, et al. Estimating the probability of neonatal early-onset infection on the basis of maternal risk factors. Pediatrics 2011;128(5):e1155-63. DOI: 10.1542/peds.2010-3464.
2. Kuzniewicz MW, Walsh EM, Li S, Fischer A, Escobar GJ. Development and Implementation of an Early-Onset Sepsis Calculator to Guide Antibiotic Management in Late Preterm and Term Neonates. Jt Comm J Qual Patient Saf 2016;42(5):232-9. DOI: 10.1016/s1553-7250(16)42030-1.